reims10_part2
Antoine JARDIN
27 novembre 2015
Set up
library(ggplot2)
df <- read.csv("~/r10.csv")
summary(df)## age educ competition
## Min. :16.00 7 :4801 competition is good:4639
## 1st Qu.:28.00 University level:4016 5 :3536
## Median :40.00 5 :3336 3 :3459
## Mean :42.88 3 :2588 2 :2960
## 3rd Qu.:56.00 6 :2501 4 :2941
## Max. :99.00 8 :2380 6 :1866
## (Other) :3922 (Other) :4143
## lrscale country sex
## 5 :5971 710. South Africa : 3033 1. Male :11948
## 6 :3303 840. United States: 2112 2. Female:11596
## 7 :2752 566. Nigeria : 1759
## 8 :2593 392. Japan : 1591
## 4 :2093 528. Netherlands : 1590
## 3 :1813 288. Ghana : 1552
## (Other):5019 (Other) :11907
## secular emancipative
## Min. :0.0000 Min. :0.0000
## 1st Qu.:0.2631 1st Qu.:0.3292
## Median :0.3878 Median :0.4567
## Mean :0.3884 Mean :0.4637
## 3rd Qu.:0.4989 3rd Qu.:0.5885
## Max. :0.9583 Max. :1.0000
## Simple Linear Regression
fm1 <- lm(emancipative ~ secular, df)
summary(fm1)##
## Call:
## lm(formula = emancipative ~ secular, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51454 -0.12723 -0.00891 0.11825 0.61822
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.316243 0.002874 110.03 <2e-16 ***
## secular 0.379542 0.006788 55.91 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1755 on 23542 degrees of freedom
## Multiple R-squared: 0.1172, Adjusted R-squared: 0.1172
## F-statistic: 3126 on 1 and 23542 DF, p-value: < 2.2e-16plot(fm1)
df$y1 <- fitted(fm1)
p1 <- ggplot(df, aes(x=secular, y=y1))
p1 + geom_point()
Multiple Linear Regression
fm2 <- lm(emancipative ~ secular + country, df)
summary(fm2)##
## Call:
## lm(formula = emancipative ~ secular + country, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51611 -0.09270 -0.00235 0.09264 0.54758
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.242028 0.006348 38.129 < 2e-16 ***
## secular 0.231710 0.005890 39.340 < 2e-16 ***
## country233. Estonia 0.130448 0.007169 18.195 < 2e-16 ***
## country288. Ghana -0.017860 0.006977 -2.560 0.0105 *
## country32. Argentina 0.195212 0.007709 25.321 < 2e-16 ***
## country356. India 0.062386 0.006942 8.986 < 2e-16 ***
## country368. Iraq -0.055660 0.007436 -7.485 7.40e-14 ***
## country36. Australia 0.270607 0.007387 36.631 < 2e-16 ***
## country392. Japan 0.168849 0.006925 24.382 < 2e-16 ***
## country528. Netherlands 0.277095 0.006918 40.052 < 2e-16 ***
## country566. Nigeria -0.030981 0.006829 -4.537 5.74e-06 ***
## country608. Philippines 0.085132 0.007191 11.838 < 2e-16 ***
## country643. Russia 0.041732 0.007063 5.908 3.50e-09 ***
## country710. South Africa 0.129192 0.006495 19.892 < 2e-16 ***
## country724. Spain 0.228535 0.007404 30.866 < 2e-16 ***
## country752. Sweden 0.379760 0.007325 51.845 < 2e-16 ***
## country76. Brazil 0.161053 0.007246 22.227 < 2e-16 ***
## country840. United States 0.217665 0.006692 32.529 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1371 on 23526 degrees of freedom
## Multiple R-squared: 0.4617, Adjusted R-squared: 0.4613
## F-statistic: 1187 on 17 and 23526 DF, p-value: < 2.2e-16plot(fm2)
df$y2 <- fitted(fm2)
p2 <- ggplot(df, aes(x=secular, y=y2))
p2 + geom_point()
p2 + geom_point() + facet_wrap( ~ country)
Multiple Linear Regression + Interaction
fm3 <- lm(emancipative ~ secular * country, df)
summary(fm3)##
## Call:
## lm(formula = emancipative ~ secular * country, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.55173 -0.09208 -0.00251 0.09288 0.55638
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.236344 0.015978 14.791 < 2e-16
## secular 0.247619 0.041491 5.968 2.44e-09
## country233. Estonia 0.197564 0.020236 9.763 < 2e-16
## country288. Ghana 0.009297 0.017406 0.534 0.59325
## country32. Argentina 0.259770 0.021150 12.282 < 2e-16
## country356. India 0.058699 0.019298 3.042 0.00235
## country368. Iraq -0.052684 0.019412 -2.714 0.00665
## country36. Australia 0.227704 0.019330 11.780 < 2e-16
## country392. Japan 0.156455 0.020103 7.783 7.39e-15
## country528. Netherlands 0.242304 0.018980 12.766 < 2e-16
## country566. Nigeria -0.042018 0.018078 -2.324 0.02012
## country608. Philippines 0.089332 0.019236 4.644 3.44e-06
## country643. Russia 0.082193 0.019967 4.117 3.86e-05
## country710. South Africa 0.165548 0.017365 9.533 < 2e-16
## country724. Spain 0.182106 0.019815 9.190 < 2e-16
## country752. Sweden 0.414020 0.021253 19.480 < 2e-16
## country76. Brazil 0.200205 0.018642 10.740 < 2e-16
## country840. United States 0.186393 0.017186 10.846 < 2e-16
## secular:country233. Estonia -0.145343 0.048372 -3.005 0.00266
## secular:country288. Ghana -0.115710 0.049916 -2.318 0.02045
## secular:country32. Argentina -0.150216 0.050994 -2.946 0.00322
## secular:country356. India 0.006459 0.048163 0.134 0.89332
## secular:country368. Iraq -0.007781 0.051411 -0.151 0.87970
## secular:country36. Australia 0.109716 0.048876 2.245 0.02479
## secular:country392. Japan 0.024503 0.049058 0.499 0.61745
## secular:country528. Netherlands 0.078307 0.047188 1.659 0.09703
## secular:country566. Nigeria 0.038257 0.048586 0.787 0.43105
## secular:country608. Philippines -0.011295 0.051753 -0.218 0.82724
## secular:country643. Russia -0.087031 0.047597 -1.829 0.06749
## secular:country710. South Africa -0.089641 0.044195 -2.028 0.04254
## secular:country724. Spain 0.105256 0.048615 2.165 0.03039
## secular:country752. Sweden -0.074005 0.049629 -1.491 0.13593
## secular:country76. Brazil -0.116595 0.049061 -2.377 0.01749
## secular:country840. United States 0.093912 0.044690 2.101 0.03562
##
## (Intercept) ***
## secular ***
## country233. Estonia ***
## country288. Ghana
## country32. Argentina ***
## country356. India **
## country368. Iraq **
## country36. Australia ***
## country392. Japan ***
## country528. Netherlands ***
## country566. Nigeria *
## country608. Philippines ***
## country643. Russia ***
## country710. South Africa ***
## country724. Spain ***
## country752. Sweden ***
## country76. Brazil ***
## country840. United States ***
## secular:country233. Estonia **
## secular:country288. Ghana *
## secular:country32. Argentina **
## secular:country356. India
## secular:country368. Iraq
## secular:country36. Australia *
## secular:country392. Japan
## secular:country528. Netherlands .
## secular:country566. Nigeria
## secular:country608. Philippines
## secular:country643. Russia .
## secular:country710. South Africa *
## secular:country724. Spain *
## secular:country752. Sweden
## secular:country76. Brazil *
## secular:country840. United States *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1365 on 23510 degrees of freedom
## Multiple R-squared: 0.4669, Adjusted R-squared: 0.4661
## F-statistic: 623.8 on 33 and 23510 DF, p-value: < 2.2e-16plot(fm3)
df$y3 <- fitted(fm3)
p3 <- ggplot(df, aes(x=secular, y=y3))
p3 + geom_point()
p3 + geom_point() + facet_wrap( ~ country)